Trade and Quote (TAQ) Data

TAQ (Trade and Quote) data is the consolidated tape of every executed trade and every posted quote across U.S. equity exchanges, published by the exchanges and distributed through vendors, forming the finest-grained public record of market activity available to researchers. Each trade record carries a timestamp, price, size, and exchange identifier; each quote record carries the best bid and ask prices and sizes at that exchange at that instant. Together they let a researcher reconstruct, second by second (or with modern feeds, nanosecond by nanosecond), exactly what the market looked like and what happened.

TAQ data is the standard input for market microstructure research — measuring bid-ask spreads, estimating price impact of trades, studying order flow toxicity, or reconstructing the National Best Bid and Offer (NBBO) by combining quotes across every exchange trading a stock. It's also used to build realistic transaction cost models: a backtest that only has daily closing prices can't tell you what it actually costs to execute a large order, but TAQ data can, because it shows the actual depth and spread available at each moment.

Working with TAQ data is a data-engineering challenge in its own right — a single liquid stock can generate millions of quote updates a day, timestamps must be carefully synchronized across exchanges (a subtlety often overlooked, since different venues' clocks can drift by milliseconds), and trades must be matched to the prevailing quote at the time to classify them as buyer- or seller-initiated.